Survey on Joint Compression and Fine-Tuning of Large Language Models: Methods, Toolchains, and Open Challenges under Resource Constraints
Article 134332 | doi: 10.1016/j.neucom.2026.134332
Language in
the real world.
Robust language understanding, evidence-aware evaluation, and the systems that connect them.
01 / Publications
Google Scholar ↗BitAbuse: A Dataset of Visually Perturbed Texts for Defending Phishing Attacks
New Mexico, USA | April 29 - May 4, 2025
OCR-Assisted Masked BERT for Homoglyph Restoration towards Multiple Phishing Text Downstream Tasks
Vol. 85, No. 3, pp. 4977-4993 | doi: 10.32604/cmc.2025.068156
GoodGPT: Counseling-Chat
Las Vegas, USA | January 11 - 14, 2025 | doi: 10.1109/ICCE63647.2025.10930112
An Efficient Fine-Tuning of Generative Language Model for Aspect-Based Sentiment Analysis
Las Vegas, USA | January 5 - 8, 2024 | doi: 10.1109/ICCE59016.2024.10444216
Scaling Up: Analyzing Classification Outcomes in Large Korean Legal Document Datasets
Vol. 15, No. 2, pp. 527-541 | doi: 10.21860/j.15.2.12
An Efficient Neural Network based on Early Compression of Sparse CT Slice Images
pp. 1-5 | doi: 10.1109/PlatCon53246.2021.9680749
02 / Datasets
BitAbuse
Visually perturbed text for studying phishing defense and text restoration, bringing real-world cases into robustness evaluation.
Explore dataset ↗03 / Research threads
Robust language understanding
Understanding and restoring text under visual perturbations, adversarial inputs, and real-world noise.
BitAbuse ↗Evaluation that supports reliable conclusions
My research interests include the gap between synthetic benchmarks and real-world conditions, and how evaluation choices shape the conclusions we draw from model outputs.
Restoration research ↗Inspectable research systems
Research infrastructure with explicit review, evidence tracking, and resumable execution.
Explore systems ↗